cronexpression.tools
DevOps

Cron Expressions for Kubernetes CronJob

Kubernetes CronJob uses standard 5-field cron syntax to schedule batch jobs. Since Kubernetes 1.27, you can specify a timezone with spec.timeZone. CronJobs create Job objects that run to completion.

Format
5-field cron
Fields
5 fields: minute hour day-of-month month day-of-week
Timezone
Cluster timezone (override with spec.timeZone in k8s 1.27+)
Validate your Kubernetes cron expression instantly
Paste any cron expression → get a plain-English explanation, next run times, and lint warnings.
Try Translator →Debugger

Common Kubernetes Cron Patterns

Every 5 minutes
*/5 * * * *

Runs a job every 5 minutes

Use caseFrequent cache warm-up, fast polling
Every hour
0 * * * *

Runs at the top of every hour

Use caseHourly data aggregation, metrics rollup
Daily database backup
0 2 * * *

Runs at 02:00 every night

Use caseNightly database dump, backup to object storage
Weekly cleanup
0 0 * * 0

Runs every Sunday at midnight

Use caseWeekly log rotation, stale data pruning
Monthly report
0 9 1 * *

Runs at 09:00 on the 1st of every month

Use caseMonthly billing export, SLA reports
Every 6 hours
0 */6 * * *

Runs four times per day

Use casePeriodic cache refresh, data sync
Weekday mornings
30 7 * * 1-5

Runs at 07:30 Monday through Friday

Use casePre-business-hours warm-up, morning digest
End of quarter
0 0 1 1,4,7,10 *

Runs at midnight on January 1, April 1, July 1, October 1

Use caseQuarterly data snapshots, financial reports

Kubernetes CronJob Manifest

yamlKubernetes
apiVersion: batch/v1
kind: CronJob
metadata:
  name: database-backup
  namespace: production
spec:
  # Run every day at 02:00
  schedule: "0 2 * * *"

  # Optional: specify timezone (requires k8s 1.27+)
  timeZone: "America/New_York"

  # Prevent overlapping runs
  concurrencyPolicy: Forbid

  # Allow up to 30s late start before skipping
  startingDeadlineSeconds: 30

  # Keep last 3 successful and 1 failed job
  successfulJobsHistoryLimit: 3
  failedJobsHistoryLimit: 1

  jobTemplate:
    spec:
      template:
        spec:
          restartPolicy: OnFailure
          containers:
            - name: backup
              image: your-backup-image:latest
              command: ["/bin/sh", "-c", "pg_dump $DB_URL > backup.sql"]

How Kubernetes Differs from Standard Crontab

Kubernetes CronJob uses standard 5-field cron syntax identical to crontab. The key addition is spec.timeZone (k8s 1.27+) for explicit timezone control, and CronJob-specific fields like concurrencyPolicy and startingDeadlineSeconds.

Common Gotchas & Pitfalls

1

Timezone defaults to the cluster's system timezone — use spec.timeZone (k8s 1.27+) for explicit control.

2

Set concurrencyPolicy: Forbid if your job must not overlap; Allow is the default.

3

startingDeadlineSeconds defines how late a missed job can start — set it to prevent pile-ups after downtime.

4

CronJob controller checks every 10 seconds, so minimum meaningful interval is around 1 minute.

5

If more than 100 schedules are missed (e.g., cluster was down), Kubernetes will not backfill them.

6

Job pods remain after completion — use successfulJobsHistoryLimit and failedJobsHistoryLimit to control cleanup.

7

Daylight saving time transitions can cause jobs to be skipped or doubled if timezone is set.

Other Platform Guides

GitHub ActionsAWS EventBridgeGCP SchedulerAzure Functions

More Cron Expression Tools

Translate, build, test, and reference cron expressions — all in one place.

TranslatorVisual BuilderTester / DebuggerCheat Sheet